The Language You Speak Predicts Your Ability To Remember The Different Parts Of Lists

GettyImages-925066994.jpgBy Matthew Warren

For decades, linguists have debated the extent to which language influences the way we think. While the more extreme theories that language determines what we can and can’t think about have fallen out of favour, there is still considerable evidence that the languages we speak shape the way we see the world in more subtle ways. 

For instance, people are better at perceiving the difference between light and dark blue if they have dedicated words for those colours (like in Russian) than if they don’t (like in English). But it turns out it’s not just the words that we use: the way in which a language is structured – its syntax – is also important. In a recent study in Scientific Reports, Federica Amici and colleagues show that the word order of a language predicts how good its speakers are at remembering the first or last parts of a list.

The team conducted memory tasks with speakers of eight different languages around the world. Four of the languages were what’s known as right-branching – Ndonga, Khmer, Thai and Italian – and four were left-branching – Japanese, Korean, Khoekhoe and Sidaama. They included 24-30 speakers of each language, and tested them in their local area, with the help of a translator where necessary.

In a right-branching language, the most important part of a phrase, known as the “head”, comes first, followed by words that give more information about that word. So in the phrase “The tiger who came to tea”, the head ‘tiger’ is at the beginning of the sentence, followed by words that reveal more about him (i.e. that he came to tea). 

 A left-branching language is the opposite: the extra information comes first, and the head is at the end of the sentence. The same phrase in a left-branching language would be something like “who came to tea, the tiger”. (English is mainly right-branching, but has some left-branching tendencies: we put adjectives before the noun that they modify, as in “the hungry tiger”, rather than “the tiger hungry”).

In right-branching languages, the meaning of a sentence is clear from the start because the head is encountered immediately. But in left-branching languages, speakers need to keep a lot of information in mind before they get to the important part of the sentence that clarifies what it’s all about. So the researchers wondered whether speakers of left-branching languages would be generally better at remembering the early parts of lists of stimuli. 

The participants completed three tests designed to tap into working memory, which involves holding information in mind while also processing other information – a key skill in language. For each test, they viewed a sequence of pictures (or some other visual sequence) while simultaneously completing a second distracting task like performing mental arithmetic or judging whether an image was symmetrical. At the end of each series they had to recall the sequence they’d seen in the correct order. These sequences varied across tests and included pictures of animals and objects, different numbers of blue circles, and red squares in a series of grids. 

The team then compared how well the speakers of the two kinds of language remembered stimuli from the first and second halves of the sequences. As they predicted, left-branching speakers were better at recalling the first half of lists than right-branching speakers, and worse at recalling the second half. 

The results show how the languages we speak can have fairly broad influences on cognition, say the authors. “Specific characteristics of a language appear to predict not only the way we perceive and conceptualize the world,” they write, “but also the way we process, store and retrieve information.”

The research also demonstrates how differences in culture and language can bias findings in psychology. For example, based almost entirely on Western participants, memory researchers have found that people tend to recall the first and last parts of lists better than the middle parts – phenomena referred to as the “primacy” and “recency” effects – which have long been considered a fundamental aspect of how memory works. It is clear from this new study that the language participants speak could have a significant impact on these kinds of results, potentially challenging whether human memory always works that way. 

But for now, the study remains preliminary. The overall differences in memory between left- and right-branching languages appear fairly small, and the eight languages the authors studied are just a tiny fraction of the more than 7,000 spoken around the world. Whether the results will be replicated in a larger sample remains to be seen. 

The word order of languages predicts native speakers’ working memory

Matthew Warren (@MattbWarren) is Staff Writer at BPS Research Digest

3 thoughts on “The Language You Speak Predicts Your Ability To Remember The Different Parts Of Lists”

  1. Intriguing report. Two unrelated comments: 1) I wonder if anyone has posited or found that LB speakers tell jokes better? I have a sense that a lot of the punch in a joke comes from having an unexpected twist at the end. Is this easier to accomplish in an LB language, or is it harder, since listeners are used to waiting for the subject at the end? and 2) I am learning Irish (which is one of those rare VSO languages BTW). I also have a new granddaughter, so I have been buying children’s books in Irish, and the first one I bought was An Tíogar a Thainig Chun Tae, or The Tiger who Came for Tea. Do you know this book, in English or Irish, or was it just a lovely serendipitous coincidence for me that you chose that phrase as your illustration?

  2. This reminds me. In the olden days of LISP and good-old-fashioned AI (1970s and 1980s), we’d write (PLUS X Y) (Lisp) and (PTRANS BLOCK1 POSITION2) for “physically move block 1 to position 2”. A Japanese postdoc (visiting the AI group at Yale) implemented a whole system using the reverse order: (BLOCK1 POSITION2 PTRANS). It took me waaaaaaaaaay longer than it should have to realize that that was the Japanese language word order. (Well, this could turn into a long story, since Japanese is case marked, and doesn’t really have word order, other than the verb being last. But wo ni is pretty natural (“wo” and “ni” are the case markers).)

    Despite being reasonably fluent in Japanese at the time, just looking at his work gave me a headache*. His code, was, of course, in ordinary Lisp order, and he didn’t seem to have any trouble dealing with that. But for his data, he just found it easier to do it in Japanese order.

    *: It didn’t help that the automatic formatting in Lisp does the wrong thing for this data order.

  3. Thank you for sharing !
    I was wondering, though, were the participants monoligual ?

Comments are closed.